1 / 31

COBOL 2002 – The Good, the Bad, and the UGLY

COBOL 2002 – The Good, the Bad, and the UGLY. Tom Ross (Based on an original presentation created by Bill Klein) Session: 8204 August 25, 2005 – 11:00 a.m. Introduction.

emele
Download Presentation

COBOL 2002 – The Good, the Bad, and the UGLY

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. COBOL 2002 – The Good, the Bad, and the UGLY Tom Ross (Based on an original presentation created by Bill Klein) Session: 8204 August 25, 2005 – 11:00 a.m.

  2. Introduction This sessions was originally created by Bill Klein and reflects his input and options; not (necessarily) those of Tom Ross or IBM. Who is Bill Klein and what does he have to do with this and other Standards? I have: • Worked on the J4 (formerly X3J4) COBOL Committee (initially representing a compiler vendor, then representing myself as an individual) • Worked on the CODASYL COBOL Committee (when it still existed) • Worked for an IBM partner on the design (and “long-dead”) implementation of SAA COBOL L2 (and the never seen L3) • Worked with GUIDE (and now SHARE) to help them understand what has, is, and might be happening with ANSI and ISO COBOL Standards SHARE Bostong, Session 8204

  3. As far as ISO 1989:2002 (the ISO 2002 COBOL Standard) goes I have: • I worked on it during the first two-thirds (or so) of its development • I really liked (and still like) many if not all of the new and enhanced features • By the time of the final public review period, I came to believe (and expressed myself rather vocally) that the final product was NOT good for the COBOL programming community (and whether or not it is good for compiler vendors may or may not depend upon whether a fully conforming compiler ever becomes available) SHARE Bostong, Session 8204

  4. As far as ISO 1989:2002 (the ISO 2002 COBOL Standard) goes (continued) • To this day, I don’t know whether I think it will or won’t benefit COBOL programmers, vendors, or others. • I don’t really know if ANY vendor will ever produce a “fully conforming” 2002 conforming compiler – much less if any programmer will ever use such a product when/if it becomes available.  (I certainly doubt that there will ever be any “certification” tests for it – so that “users” can TELL if a compiler is or is not truly conforming.) SHARE Bostong, Session 8204

  5. What is the status of the COBOL Standard? • The ISO 2002 Standard was “approved” in December of 2002 • The ’85 Standard (ANSI and ISO) – with or without its two Amendments - is no longer an officially recognized Standard • The first TR (Technical Report) for “Object Finalization” has been completed and approved • Two more TR’s (for native XML support and for Collection Classes) are progressing and should be finished and approved within the next year or two (or so) • The NEXT ISO COBOL Standard is currently scheduled to be finished and approved in 2008 SHARE Bostong, Session 8204

  6. What is the status of the COBOL Standard? (continued) • For details of the ongoing (and future) revision work, see http://www.cobolportal.com/j4/files/05-0108.doc The first draft of this next Standard should be available for “public review” in the middle of next year (or so). • There is (as far as I know) currently no FIPS (US government) COBOL (or other programming language – except maybe ADA and SQL) Standard – much less government recognized validation suite for COBOL. SHARE Bostong, Session 8204

  7. Processor-dependent language elements To meet the requirements of standard COBOL, the implementor shall document the processor-dependent language elements for which the implementation claims support. Language elements that pertain to specific processor-dependent elements for which support is not claimed need not be implemented. The decision of whether to claim support for a processor-dependent language element is within an implementor's discretion. Factors that may be considered include, but are not limited to, hardware capability, software capability, and market positioning of the processor. SHARE Bostong, Session 8204

  8. IBM’s implementation of “Substantive Changes” – What’s New With Enterprise COBOL V3R4, IBM introduced extended support for National (Unicode compatible UTF-16) Data types – implemented in accordance with the specification in ISO 2002 COBOL. As indicated in the LRM “Summary of Changes”: Support for national (Unicode UTF-16) data has been enhanced. Several additional kinds of data items can now be described implicitly or explicitly as USAGE NATIONAL: • External decimal (national decimal) items • External floating-point (national floating-point) items • Numeric-edited items • National-edited items • Group (national group) items, supported by the GROUP-USAGE NATIONAL clause SHARE Bostong, Session 8204

  9. Summary IBM’s implementation of “Substantive Changes” • Already Implemented – 56 • Comparable Functionality is available in Enterprise COBOL – 6 • Items Partially implemented in Enterprise COBOL - 5 • Items not yet implemented in Enterprise COBOL – 88 NOTE: this count has not been updated since the latest requirements were submitted to IBM and since IBM submitted their latest responses. SHARE Bostong, Session 8204

  10. Some detailed issues (good and bad) • Reference format changes • GOOD: In-Line Comments 05 Field-X Pic XX *> Used in calculating the current THINGY . . . MOVE ABC to XYZ *> Current-XYZ LMN *> Saved XYZ. • GOOD: New Style Literal continuation Move ‘ABC’- ‘XYZ’ to 6-bytes • WEIRD: Identification Division header is optional SHARE Bostong, Session 8204

  11. Some detailed issues (good and bad) (cont) • BAD: A-/B-margin removal Procedure Division. Para1. Move ABC to zZz. XYZ Section. CALL “ABC”- “DEF” • BAD: 255 character (not byte) lines Move “ABCD杏杳枉89012345678921234567893123456789412345678951234567896123456789712345678981234567899123456789a123...” to Recv-Field • BAD: potential problems with loss of columns 73- 80 information SHARE Bostong, Session 8204

  12. Performance issues • Standard arithmetic NOTE: For an introduction to what “Standard Arithmetic” is and does, see: Appendix E: Introduction to Standard arithmetic • Common Exception handling NOTE: For an introduction to what “Common Exception Handling” is and does, see: Appendix D: Introduction to Common exception processing • Table SORT • CALL prototypes and compile-time parameter checking SHARE Bostong, Session 8204

  13. Detailed "oddities" and miscellaneous • GOOD: Various new Standard-defined data types • 4 sizes of “true binary” (cf. TRUNC(BIN)) data USAGEs • 3 sizes of floating point items • Pointer and Procedure-Pointer types (already available with Enterprise COBOL) • Bit and Boolean data items • BAD: IF NUMERIC on BINARY items (against picture clause) 05 Group-1. 10 Binary-1 Pic S9(04) Binary. . . . Move High-Values to Group-1 If Binary-1 Numeric Display “Not true for ISO 2002” SHARE Bostong, Session 8204

  14. BAD: Lots of new reserved words For Example: ADDRESS • ALIGNED • ALLOCATE • AS • BIT • COLUMNS • CONDITION • CONSTANT • CURSOR • DEFAULT • FACTORY • FORMAT • FREE • GET • INHERITS • LOCALE • METHOD • MINUS NESTED • OPTIONS • OVERRIDE PRESENT • PROPERTY • PROTOTYPE • RAISE • RAISING • REPOSITORY • RESUME • RETRY • RETURNING • SCREEN • SELF • SHARING • SOURCES • SUPER • UNIVERSAL • UNLOCK • VALID • VALIDATE Detailed "oddities" and miscellaneous (cont) SHARE Bostong, Session 8204

  15. Detailed "oddities" and miscellaneous (continued) • WHO KNOWS: the "report writer is required - but may require add-on product" question For a useful (interesting) paper on Report Writer and the ’02 Standard – including LOTS of examples, see: http://www.cobolstandard.info/rw.htm • GOOD (MAYBE): ALL customers will be getting AND paying for "internationalization" support in the compiler Check out what IBM introduced as one area of major enhancement in Enterprise COBOL V3R4 – i.e. NATIONAL numeric support BAD, BUT: ALLOCATE and FREE and "BASED" phrase vs current IBM support for SET ADDRESS OF linkage-section-items SHARE Bostong, Session 8204

  16. Detailed "oddities" and miscellaneous (continued) • ALSO SCARY, BUT USEFUL: Bit twiddling 01 My-flag PIC 1111 USAGE BIT VALUE B"0000". 01 My-flag-2 PIC 1111 USAGE BIT. ... MOVE B"0011" to My-flag-2 *> Initialize My-flag-2 ... *> set the bits in My-flag to the *> reverse of My-flag-2, 1100. COMPUTE My-flag = B-NOT My-flag-2 ... *> turn off all the bits in My-flag. COMPUTE My-flag = My-flag B-AND B"0000" ... *> set bit 1 ON in My-flag-2, keeping other bits unchanged. COMPUTE My-flag-2 = My-flag-2 B-OR BX"8" ... *> Alternatively, the last COMPUTE statement could be replaced by: *> set bit 1 ON using reference*> modification MOVE B"1" to My-flag-2(1:1) SHARE Bostong, Session 8204

  17. Detailed "oddities" and miscellaneous (continued) • DOES ANYONE CARE?: full (Java and C++ "independent") OO The ISO 2002 OO definition is COBOL-based and should (as far as I can tell) work in • A COBOL-only OO environment • A mixed COBOL and C++ (or C#) environment • A mixed COBOL and Java environment • GOOD, BUT: VALIDATE NOTE: For an introduction to what the “VALIDATE” facility is and does, see Appendix C: Introduction to the Validate facility • TRIVIAL, BUT: WRITE from Literal Write Some-Record from Spaces Write Other-Record from All “X” SHARE Bostong, Session 8204

  18. NICE BUT: • Numeric items may be defined up to 31 digits (Available with ARITH(EXTEND) compiler option in Enterprise COBOL – for some, but not all, data types) • FORMAT clause can do (implicit) encoding conversions upon input or output of file I/O • Implementor defined successful file status codes (cf. “97” and VSAM could now be “0x”) • Inline PERFORMs may have an AFTER phrase • Non-floating currency symbol can appear at the end of a PICTURE clause SHARE Bostong, Session 8204

  19. NICE BUT: (continued) • READ PREVIOUS and START LESS THAN • Recursion and Local-Storage section (in both OO and procedural code) • REPLACE ALSO statement now may be “cumulative” rather than simply canceling and replacing previous REPLACE statements • SET condition-name TO FALSE (and WHEN FALSE clause) • “Split” RECORD and ALTERNATE RECORD keys (keys made up of multiple non-contiguous fields) SHARE Bostong, Session 8204

  20. NICE BUT: (continued) • STOP WITH STATUS (possible replacement for either RETURN-CODE or CEE3ABD) • “Strong typing” (via TYPEDEF and TYPE attribute) Similar but not the identical to the new SAME clause (cf. COBOL/400 “LIKE” clause) • Subscripting using arithmetic expressions • Underscore in user-defined words (NOT treated as equivalent to a hyphen) • User-defined functions SHARE Bostong, Session 8204

  21. New (and better?) Ways of doing Old tasks • Accept, Display, and Screen Section • Concatenation Expressions Move “IBM” & “ Z-literal Extension” & “ isn” & ‘’’’ & “t part of 2002 Standard” & Low-Values To result-field • Conditional Compilation >>IF PROD-TEST = “TEST” Display “Some Var:” Some-Variable Perform Only-in-Test *> the following evaluation is done at compile-time, not run-time >>If Too-Many-Errors > Error-Cnt Perform Really-Bad >>Else >>DEFINE Error-Cnt AS Error-Cnt + 1 >>END-IF >>ELSE Perform Update-Production-Files >>END-IF SHARE Bostong, Session 8204

  22. New (and better?) Ways of doing Old tasks (cont) • Constants Working-Storage Section. 01 Start-of-WS Constant Value “Working-Storage for Program: ABC Starts Here”. • COPY and REPLACE “partial word” replacement COPYMEM *> source code of copy member is: 01 Group-Item 05 GI-Field-1 Pic X. 05 G1-Field-2 Binary-Char Signed. *> COPY statement in main program is Copy COPYMEM Replacing leading ==GI-== by ==WS-GI-== Trailing ==-2== by ==-Single-byte-binary== . • File Sharing and Record Locking (controlled in source code) SHARE Bostong, Session 8204

  23. New (and better?) Ways of doing Old tasks (continued) • HIGHEST-/LOWEST Algebraic Intrinsic Functions Repository. Function All Intrinsic. . . . 01 Num-Fields. 05 N1 Pic S9(04)V99 Packed-Decimal. *> not even number of digits 05 N2 Float-Short. 05 N3 Pic $,$$9,999.99 . . . Move Highest-Algebraic to N1 If N2 = Lowest-Algebraic Move lowest-algebraic to N3 • INITIALIZE with new options 01 Group-Item. 05 Value “Implicit Filler”. 05 Named-National Value NX”AB12CD98”. 05 Filler Pic 1 Value B”1”. 05 Var-Tabl. 10 Each-Elem Occurs 10 times Pic 9 Value 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, (1) to (10). . . . Initialize Group-Item With Filler All to Value . SHARE Bostong, Session 8204

  24. New (and better?) Ways of doing Old tasks (continued) • Reversible order for conditional phrases Compute ABC = XYZ + 1 Not On size error Perform No-Truncation On Size Error Perform Too-big-to-handle End-Compute • Select/Assign USING (for source code file specification) Select ABC Assign Using Inp-Name. Select XYZ Assign Using Out-Name. . . . 01 Misc. 05 Inp-Name 10 Value “DDNAME=” 10 Dynamic-Name Pic X(8) 05 Out-Name Value “DSN=HLQ.NEWQSAM,DISP=(NEW,CATLG,DELETE),”- “UNIT=SYSDA,DCB=(RECFM=F,LRECL=80),”- “SPACE=(CYL(2,1))” . . . If Week-Day Move “WEEKDDN” to Dynamic-Name *> must have WEEKDN in JCL as DDNAME Else Move “NOTWORK” to Dynamic-Name End-IF Open Input ABC Output XYZ SHARE Bostong, Session 8204

  25. New (and better?) Ways of doing Old tasks (continued) • VALUE clauses for individual Table entries 05 Level-1 Occurs 10 times. 10 Level-2 Occurs 3 times PIC X Value “A” “B” “C” From (2 2) to (3 1) ”X” “Y” “Z” From (1 1) to (1 3) Space from (3 2). SHARE Bostong, Session 8204

  26. Web-Pages of (possible) Interest For additional information on any of the documents or sites listed below or to get on the “notification” list for public reviews of future Technical Reports (TR’s) and the next ISO COBOL Standard (currently expected in 2009), send an email to the J4 chair at: Don.Schricker@microfocus.com • To obtain a downloadable PDF copy of the ANSI version of ISO 1989:2002 Standard (for a minimal price), see: http://www.techstreet.com/cgi-bin/basket?action=add&item_id=3574917 • To obtain a printed copy of the ANSI version of the ISO 1989:2002 Standard (for a still minimal price), see: http://www.techstreet.com/cgi-bin/basket?action=add&item_id=3574918 SHARE Bostong, Session 8204

  27. Web-Pages of (possible) Interest (continued) • To obtain a downloadable or CD-ROM version of the ISO version of ISO 1989:2002 Standard (for a not so reasonable price), see: http://www.iso.org/iso/en/CatalogueDetailPage.CatalogueDetail?CSNUMBER=28805 • For a copy of the (Dec 02, 2004) “Technical Corrigendum 1” to ISO 1989:2002, see: http://www.cobolportal.com/j4/files/04-0216.doc • For a copy of the Draft (Feb 4, 2005) “Technical Corrigendum 2” to ISO 1989:2002, see: http://www.cobolportal.com/j4/files/05-0016.doc SHARE Bostong, Session 8204

  28. Web-Pages of (possible) Interest (continued) • For a copy of the (final draft) of the (now approved) Technical Report on “Object finalization for programming language COBOL”, see: http://www.cobolportal.com/j4/files/03-0046.doc • For a current (Aug 18, 2005) draft of the “Native COBOL Syntax for XML Support”, see: http://www.cobolportal.com/j4/files/05-0172.doc • For a current (as of Feb 11, 2005) draft of the “Collection Class Library for programming language COBOL”, see: http://www.cobolportal.com/j4/files/05-0042.doc SHARE Bostong, Session 8204

  29. Web-Pages of (possible) Interest (continued) • For “COBOL Standardization Process – Status,” see: http://www.cobolstandards.com/ • For access to ALL of the current (and recent past) documents being (or having been) processed by J4, see: http://www.cobolportal.com/j4/index.asp?bhcp=1 • For (international) WG4 status, see: http://www.cobolstandard.info/wg4/wg4.html SHARE Bostong, Session 8204

  30. In Conclusion Even if every current customer of IBM’s were to ask IBM to implement every feature, change, and enhancement in the ISO 2002 COBOL Standard, I don’t know if/when IBM would do so. However, as this presentation has demonstrated, IBM has already met many of its customers’ identified needs by implementing some features of the new Standard while delaying implementation of other features that may or may not actually be desired or desirable. SHARE Bostong, Session 8204

  31. In Conclusion (continued) If there are specific features that are not currently implemented but that seem desirable to your installation (or as if they would meet existing or foreseen business needs), it is important that you communicate this to IBM – via SHARE requirements and your local IBM marketing branch. Similarly, although it is unlikely that IBM needs a “do NOT implement” requirement for undesirable features, it may be worth your while to communicate your views on such changes sooner than later – to avoid eventually paying for the resources need to develop features that you do not want or need. SHARE Bostong, Session 8204

More Related